Best-fit parameter values for group-averaged data.

For each parameter, we report the best-fit value and the 68% confidence interval obtained by a bootstrapping procedure. The value of σ is reported in units of excitatory drive (see Eq 2). In the row of wx, we also report the stimulus-driven attentional gain factor of the neuron tuned to the target in the no-, small-, medium-, large- and split-competitor conditions (corresponding to the five values in the parenthesis, respectively). In the row of wv, the goal-driven attentional gain factor of the neuron tuned to the target is reported too. This value is the same across conditions because the spatial spread of goal-driven attention did not change with competitor (see details in Table 1).Best-fit parameter values for group-averaged data.